DocumentCode :
1896569
Title :
Application of Support Vector Regression Algorithm in Colleges Recruiting Students Prediction
Author :
Ying, E.
Author_Institution :
Sch. of Foreign Languages, Harbin Univ. of Sci. & Technol., Harbin, China
Volume :
2
fYear :
2012
fDate :
23-25 March 2012
Firstpage :
173
Lastpage :
176
Abstract :
Support vector regression algorithm is applied to colleges recruiting students prediction in the paper. As colleges recruiting students prediction is a nonlinear regression problem, the input training data of colleges recruiting students are nonlinearly mapped into a high dimensional space in support vector regression model. The amount of colleges recruiting students of Sichuan province from 2000 to 2008 is used to prove the effectiveness of support vector regression method. Then,the forecasting curves of support vector regression method and BP neural network and the comparison of forecasting error for amount of colleges recruiting students between support vector regression method and BP neural network are given in this study.The comparison results of forecasting error for amount of colleges recruiting students between support vector regression method and BP neural network indicate that support vector regression method has a higher forecasting accuracy than BP neural network.
Keywords :
backpropagation; educational institutions; forecasting theory; further education; genetic algorithms; regression analysis; support vector machines; BP neural network; Sichuan province; colleges recruiting students prediction; forecasting curves; forecasting error; genetic algorithm; nonlinear regression problem; support vector regression algorithm; Educational institutions; Forecasting; Genetic algorithms; Prediction algorithms; Predictive models; Support vector machines; Vectors; colleges recruiting students; forecasting performance; support vector regression algorithm;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Science and Electronics Engineering (ICCSEE), 2012 International Conference on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4673-0689-8
Type :
conf
DOI :
10.1109/ICCSEE.2012.456
Filename :
6187928
Link To Document :
بازگشت